[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-120041-en":3,"doc-seo-120041-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},120041,4398048950312,"Violet","https://ap-avatar.wpscdn.com/avatar/400002538284de19e3c?_k=1778320343897328908",8,"Research & Report","GeoShapley - A Game Theory Approach to Measuring Spatial Effects in Machine Learning Models","GeoShapley presents a game-theoretic method for quantifying spatial effects in machine learning predictions. It extends the Shapley value framework by treating location as a player in a prediction game, allowing measurement of location importance and interaction synergies between location and other features. The approach is model-agnostic and works with statistical or black-box machine learning models. Its interpretation links directly to spatially varying coefficient models for spatial effects and additive models for non-spatial effects, validated via simulated data and illustrated through house price modeling with a geospatial dataset. The method is provided as an open-source Python package, geoshapley.","GeoShapley: A Game Theory Approach to Measuring Spatial Effects in Machine Learning Models  \nZiqi Li ([Ziqi.Li@fsu.edu](Ziqi.Li@fsu.edu))  \nDepartment of Geography, Florida State University, Tallahassee, United States  \nForthcoming in the Annals of the American Association ofGeographers  \nAbstract  \nThis paper introduces GeoShapley, a game theory approach to measuring spatial effects in machine learning models. GeoShapley extends the Nobel Prize-winning Shapley value framework in game theory by conceptualizing location as a player in a model prediction game, which enables the quantification of the importance of location and the synergies between location and other features in a model. GeoShapley is a model-agnostic approach and can be applied to statistical or black-box machine learning models in various structures. The interpretation of GeoShapley is directly linked with spatially varying coefficient models for explaining spatial effects and additive models for explaining non-spatial effects. Using simulated data, GeoShapley values are validated against known data-generating processes and are used for cross-comparison of seven statistical and machine learning models. An empirical example of house price modeling is used to illustrate GeoShapley's utility and interpretation with real world data. The method is available as an open-source Python package named geoshapley.  \nKeywords: Shapley, Explainable AI (XAI), GeoAI, spatial processes, non-linear, interaction, spatial effects  \n1 Introduction  \nMachine learning and AI have been increasingly used to model geospatial phenomena with promising performance across various domains. Its strength lies in the capacity to account for complex structures from vast and heterogenous datasets in a scalable, flexible, and accurate manner. Machine learning becomes the preferred choice when the primary goal is in achieving \"predictive accuracy\" in applications such as land cover classification (Camps-Valls et al., 2013; Zhang et al., 2019), weather forecast (Espeholt et al., 2022; Bi et al., 2023), traffic prediction (Derrow-Pinion et al., 2021), and object detection (Li and Hsu, 2020; Xie et al., 2020) . However, from a scientific discovery perspective, geographers are arguably more interested in seeking explanations, aiming to understand the relationships and processes underlying geospatial phenomena. To this end, the black-box nature of machine learning models begins to diminish their utility and overshadow their advantages compared to more interpretable statistical alternatives.  \nRecent advances in the field of eXplainable AI (XAI) provide a solution to explain black-box machine learning. XAI encompasses a set of processes and methods that enable human users to comprehend and trust the results generated by machine learning algorithms (Das and Rad, 2020) . Its objectives are to improve understanding of the underlying decision processes, to enhance credibility and confidence in the model parameters and outcomes, to acknowledge limits and uncertainties, and to inform future model development (Gunning et al., 2019; Murdoch et al., 2019) . While the term \"XAI\" is relatively recent (coined by the US Defense Advanced Research Projects Agency (DARPA) in 2017, Gunning and Aha, 2019), the concept of interpretable machine learning isnot novel. Interpretation techniques like permutation feature importance and partial dependence plots have seen extensive application. Permutation feature importance quantifies a feature's contribution to the overall model's predictive accuracy. Removing or altering an important feature is expected to reduce model accuracy, whereas removing an irrelevant feature should have no impact (Breiman, 2001) . The partial dependence plot illustrates the marginal effect that one or two features have on the predicted outcome of a machine learning model (Friedman, 2001) . It can reveal whether the relationship between the outcome and a feature is linear, non-linear, monotonic, or more","cbCaiqqDupwl7NZE","https://ap.wps.com/l/cbCaiqqDupwl7NZE","pdf",4283598,1,34,"English","en",105,"# Introduction\n## Explaining black-box machine learning with XAI\n## Global vs local interpretation methods\n## LIME and its local proximity challenge\n## SHAP and the Shapley value framework\n## Motivation for spatially aware explanations\n# GeoShapley concept and methodology\n## Location as a player in a prediction game\n## Quantifying location importance and feature synergies\n## Model-agnostic applicability\n# Validation and comparisons\n## Simulated data experiments\n## Cross-comparison of statistical and ML models\n# Empirical example\n## House price modeling and real-data interpretation\n# Implementation\n## Open-source Python package (geoshapley)","[{\"question\":\"What problem does GeoShapley address in machine learning for geospatial data?\",\"answer\":\"GeoShapley targets the need to measure and explain spatial effects that black-box machine learning models may not make interpretable, especially compared with traditional spatially informed statistical approaches.\"},{\"question\":\"How does GeoShapley extend the Shapley value framework?\",\"answer\":\"GeoShapley conceptualizes location as a player in a prediction game, enabling quantification of location importance and interactions (synergies) between location and other features.\"},{\"question\":\"How is GeoShapley validated and what example is used to demonstrate its usefulness?\",\"answer\":\"The method is validated using simulated data against known data-generating processes and then cross-compared across seven statistical and machine learning models; an empirical house price modeling example illustrates interpretation with real-world data.\"}]","GeoShapley - A Game Theory Approach to Measuring Spatial Effects in Machine Learning Models | PDF",1785727852,86,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"geoshapley-a-game-theory-approach-to-measuring-spatial-effects-in-machine-learning-models","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/geoshapley-a-game-theory-approach-to-measuring-spatial-effects-in-machine-learning-models/120041/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-03",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does GeoShapley address in machine learning for geospatial data?","Question",{"text":75,"@type":76},"GeoShapley targets the need to measure and explain spatial effects that black-box machine learning models may not make interpretable, especially compared with traditional spatially informed statistical approaches.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does GeoShapley extend the Shapley value framework?",{"text":80,"@type":76},"GeoShapley conceptualizes location as a player in a prediction game, enabling quantification of location importance and interactions (synergies) between location and other features.",{"name":82,"@type":73,"acceptedAnswer":83},"How is GeoShapley validated and what example is used to demonstrate its usefulness?",{"text":84,"@type":76},"The method is validated using simulated data against known data-generating processes and then cross-compared across seven statistical and machine learning models; an empirical house price modeling example illustrates interpretation with real-world data.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]